Skills Plugins MCP Prompt Model 博客 我的中心
数据分析与咨询 #data #seo #database #research

seo-keywords

Shopify keyword opportunity research sub-skill. Maps keywords from the drinkware keyword database to search intent, difficulty tier, and store coverage status. Produces an opportunity table — not scored findings. Use when user runs /seo keywords or when seo-audit delegates keyword checks.

DeepseekModel 官方收录技能 质量 良好 · 48 v1.0.0

获取

https://deepseekmodel.com/api/download.php?id=karvedconcepts-claude-seo-skills-seo-keywords-skill-md&format=skill
下载 .skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用
.skill 文件中 system_prompt 字段的实际内容。
name seo-keywords description Shopify keyword opportunity research sub-skill. Maps keywords from the drinkware keyword database to search intent, difficulty tier, and store coverage status. Produces an opportunity table — not scored findings. Use when user runs /seo keywords or when seo-audit delegates keyword checks. SEO Keyword Opportunities Keyword research sub-skill for Shopify stores. Surfaces keyword opportunities from drinkware-keywords.md , mapping each keyword to search intent (transactional/navigational), difficulty tier (High/Medium/Low), and coverage status against the store's crawl data. This is a research companion skill — it does NOT produce PASS/FAIL findings or contribute to the SEO Health Score. Quick Reference Command What it does /seo keywords <url> Run keyword opportunity analysis against a Shopify store (via seo-audit) Receives crawl data from parent orchestrator Context Intake This skill is invoked by seo-audit or directly via /seo keywords <url> . When invoked by seo-audit: Store URL and crawl data path are provided by the parent — do NOT re-ask Read seo-audit-crawl.json from the provided path Proceed directly to Execution Logic When invoked directly: If store URL not provided, read ~/.claude/skills/seo/profiles/karved.json and use preset Run python3 ~/.claude/skills/seo/scripts/crawl_store.py <url> to generate crawl data before proceeding Save output as seo-audit-crawl.json in the current directory Then proceed to Execution Logic Execution Logic Execute these steps in order: Read seo-audit-crawl.json from the provided path (or current directory if invoked standalone) Load ~/.claude/skills/seo/references/drinkware-keywords.md — this is the keyword source (required) Parse all keywords from drinkware-keywords.md, organized by their section membership For each keyword, assign Intent using the Search Intent Mapping table: drinkware-keywords.md Section Intent Primary Keywords transactional Secondary Keywords transactional Long-Tail Keywords transactional B2B Intent Keywords transactional Local Intent Keywords navigational UV Printing Variants transactional Laser Engraving Variants transactional B2B and Corporate Variants transactional Product-Specific Variants transactional Occasion and Use-Case Variants transactional Local and Proximity Variants navigational For each keyword, assign Tier (difficulty proxy) using section membership: drinkware-keywords.md Section Tier Primary Keywords High difficulty Secondary Keywords Medium difficulty Long-Tail Keywords Low difficulty B2B Intent Keywords Low difficulty Local Intent Keywords Low difficulty UV Printing Variants Low difficulty Laser Engraving Variants Low difficulty B2B and Corporate Variants Low difficulty Product-Specific Variants Low difficulty Occasion and Use-Case Variants Low difficulty Local and Proximity Variants Low difficulty For each keyword, check Coverage against crawl data: For each page in seo-audit-crawl.json : Check fields: title , h1 , meta_description (case-insensitive substring match) If keyword appears in ANY of these fields on ANY page: ✓ Covered / Gap = No If keyword appears in NONE of these fields on ANY page: ✗ Not covered / Gap = Yes If meta_description field is absent from crawl data: fall back to title + h1 only and note at the bottom of output: "Note: meta_description field not available in crawl data — coverage checked against title and H1 only." Requires meta_description field in seo-audit-crawl.json — provided by crawl_store.py (Phase 9) Build the opportunity table with all keywords Write output to seo-keywords-opportunities.md in the current directory Output Format Write findings to seo-keywords-opportunities.md in the current directory. The output file seo-keywords-opportunities.md uses this format: Header # Keyword Opportunities **Store:** {store_url} **Date:** {audit_date} **Keywords checked:** {total_count} from drinkware-keywords.md Opportunity Table Present High and Medium difficulty keywords in full. For Low difficulty keywords, show only rows where Gap = Yes (uncovered). Summarize covered Low difficulty keywords as a count. ## High Difficulty Keywords | Keyword | Intent | Tier | Current Coverage | Gap | |---------|--------|------|-----------------|-----| | custom tumblers | transactional | High difficulty | ✓ Covered | No | | promotional drinkware | transactional | High difficulty | ✗ Not covered | Yes | | ... | ... | ... | ... | ... | ## Medium Difficulty Keywords | Keyword | Intent | Tier | Current Coverage | Gap | |---------|--------|------|-----------------|-----| | insulated stainless steel tumbler | transactional | Medium difficulty | ✗ Not covered | Yes | | ... | ... | ... | ... | ... | ## Low Difficulty Keywords — Gaps Only | Keyword | Intent | Tier | Current Coverage | Gap | |---------|--------|------|-----------------|-----| | UV printing on tumblers | transactional | Low difficulty | ✗ Not covered | Yes | | ... | ... | ... | ... | ... | _{N} Low difficulty keywords are covered and not shown._ ## Summary **Total:** {N} keywords checked. {X} gaps identified ({Y} High difficulty, {Z} Medium difficulty, {W} Low difficulty). > Coverage check: keyword appears in title, H1, or meta description of any store page (case-insensitive substring match). > Tier = difficulty proxy based on section membership in drinkware-keywords.md. > Intent derived from Search Intent Mapping in drinkware-keywords.md. IMPORTANT: Do NOT include a score line. This is a research companion skill that surfaces opportunities, not scored audit findings. Audit skills (seo-technical, seo-content, etc.) end their output with a weighted score — seo-keywords-opportunities.md does not end with a score line. Quality Gates This skill inherits quality gates from seo/SKILL.md by reference. No quality gate enforcement is needed — this skill produces opportunity tables, not PASS/WARNING/FAIL findings. The quality gates apply to audit skills only. Reference Files Load these on-demand as needed — do NOT load all at startup. Path resolution: All references are installed at ~/.claude/skills/seo/references/ . references/drinkware-keywords.md — Required: keyword source with all 11 sections (Primary, Secondary, Long-Tail, B2B Intent, Local Intent, and 6 Karved expansion subsections). Section membership determines tier and intent. references/benchmarks.md — Optional: load only when additional industry context is needed for opportunity prioritization
Agent 识别该技能的关键词,点击任意一个即可复制。

该技能未提供触发词。

下载的 .skill 包内含以下字段。
字段 说明
format格式标识(skill/v1)
skill_id技能唯一 ID
name技能名称
version版本号
description技能描述
category所属分类(数组)
trigger_words触发词列表
tags标签列表
source来源标识
source_url来源链接(本页地址)
exported_at导出时间(每次下载生成)
system_prompt系统提示词正文
model_config模型参数:provider / model / temperature / max_tokens / top_p
examples示例
install_guide各平台导入说明(Coze / Dify / Claude / 自定义框架)
同一份技能可按不同平台格式导出。
.skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用 下载
.skillpro 增强格式,额外含脚本 / 工具 / 依赖 / 钩子占位 下载
.json 纯 JSON 导出,只含 system_prompt 与模型参数 下载
Coze 带 frontmatter 的 Markdown,Coze 平台导入用 下载
Dify Dify DSL,创建应用后直接导入 下载

每日精选 Skill 推荐,免费送到你邮箱

输入邮箱,每天接收一个精选 AI Agent 技能推荐。完全免费,持续更新。

提交后我们会发送一封确认邮件,点击邮件里的链接才会开始收信。

完全免费,取消任意时间。我们不会发送垃圾邮件。